This article discusses in situ metallography

Understanding the application to fire damage assessment. CS

microstructures need to be interpreted
correctly, however, to take advantage of
the capabilities of this method.

Microstructure of Constituents
One most often expects CS to be used
in hot-worked or normalized conditions,
when the microstructure consists of pearl-

Overheated ite colonies in a matrix of ferrite grains of

a given average size. Pearlite is composed
of alternating plates of cementite (Fe3C)
and ferrite. Figure 1 shows low CS pearl-

Carbon Steel
ite colonies, etched in 5% Picral, that are
lamellar in the normalized condition and
spheroidized by subcritical annealing. The
number of pearlite colonies increases as
the carbon content increases, but the
JORGE L. HAU, Capstone Engineering Services, Inc. pearlite morphology stays essentially the
same.
Metallography and in situ metallography are CS is heated to the level that it becomes
powerful tools in failure analyses, metallurgical a single-phase alloy comprising only
evaluation, fitness-for-service, and fire damage austenite grains during hot working at the
manufacturing stage. The two-phase mi-
assessment. Although carbon steel (CS) has been the crostructural constituents of CS are pro-
traditional construction material for many years, CS duced afterward by direct transformation
microstructure misinterpretation still occurs. This article of austenite to ferrite and by the eutectoid
attempts to clarify some of these, in particular the effect of transformation involving austenite, ferrite,
overheating on grain growth misinterpretations. and cementite processes, occurring at the
critical temperature of ~727°C (1,340°F).
The steel transforms to 100% austenite at
~820 to 840°C (1,510 to 1,540°F), de-
pending on the carbon content. Hot
working is carried out at 1,100°C (2,010°F)

T
he bulk of oil-refining or even higher.
equipment is built with Almost all of the rolled steels show
carbon steel (CS), and must various degrees of preferential grouping
operate in a large variety of of ferrite and pearlite in alternate bands
working conditions and aligned in the rolling direction, referred
process environments. Cor- to as banding. A common misinterpreta-
rosion engineers and metal- tion of this ferrite-pearlite banding is to
lurgists make extensive use of metallo- attribute it to rolling. When hot rolling is
graphic methods, from the reception of performed, however, the steel is pure
incoming CS to the investigation of a austenite. Ferrite and pearlite are never
failure. In situ metallography has also rolled during hot working. Pearlite band-
improved greatly, and it is currently pos- ing is thus correlated with the segregation
sible to produce acetate replicas that can of alloying elements (other than carbon)
give much better resolution compared to developed in the original ingot. This seg-
years ago, allowing examination at the regation persists even after rolling, causing
highest possible magnification in optical ferrite and pearlite to precipitate in these
microscopy or an examination of the alternate preferential bands. They can ap-
replica in a scanning electron microscope. pear, disappear, and reappear in the same
46 MATERIALS PERFORMANCE November 2004

NOV 2004 MP pp 1-51.indd 46 10/11/04 1:46:24 PM

 p g @@ f p g @@

FIGURE 1

(a) (b)

Pearlite in low CS, (a) normalized and (b) spheroidized.

steel piece, depending on the cooling rate as in Figure 1(a) but after it has been and several nucleation sites in each grain.
from austenitizing temperatures.1 heated to 700°C (1,300°F) for 1 h and The process is illustrated in Figure 2. A
cooled slowly in the furnace. Notice that real case is shown in Figure 3, taken from
Partial Transformation stress-relieving heat treatments carried the partially transformed, heat-affected
The transformation process of CS un- out at ~600°C (1,110°F) may cause some zone of a CS weld. The final microstruc-
dergoes an intermediate stage when the degree of cementite spheroidization in ture within what were pearlite colonies
structure transforms into a two-phase al- the pearlite colonies. has transformed to a finer ferrite-pearlite
loy comprising only austenite and ferrite. The grain growth process is highly de- microstructure because this double trans-
The transformation processes are revers- pendant on the amount of prior cold formation process occurs during the heat-
ible upon heating and cooling. The region working. However, because CS is nor- ing and cooling cycle. The ferrite-pearlite
where austenite and ferrite coexists is re- mally supplied without any cold working microstructure may not be that small, as
ferred to as the partial transformation or at all, grain growth does not occur when in Figure 3, if the heating is maintained
two-phase region, lying in between 727°C the steel is critically heated. As soon as the for a longer period of time. Nevertheless,
and 900°C (1,652°F). It is a function of temperature reaches 727°C, the pearlite the case serves to illustrate that refinement
the carbon content of the steel. transforms to austenite and the preexist- rather than grain growth tends to occur
Figure 2 is a schematic representation ing ferrite grains stay unaffected. Within when heating CS within the partial trans-
of the expected microstructural changes this region of partial transformation the formation region.
occurring in a CS that is heated to the amount of ferrite decreases as the heating
critical temperature of 727°C and temperature increases. With slow to mod- Total Transformation
slightly beyond. Because there is no erate cooling, this newly formed austenite The double transformation process
transformation of the steel below this transforms back into a ferrite and cement- also occurs when heating and cooling steel
temperature, the microstructure stays ite microstructure. The transformation from beyond the upper critical tempera-
essentially the same—except that the process is governed by nucleation and ture. Indeed this heating would be like
cementite within the pearlite colonies growth. Hence any single grain or colony normalizing the steel, which usually
tends to spheroidize when approaching will give rise to more than one new grain causes a finer microstructure than the
727°C. Figure 1(b) shows the same steel because there are several grain boundaries original one. During a fire this heating
November 2004 MATERIALS PERFORMANCE 47

NOV 2004 MP pp 1-51.indd 47 10/11/04 1:46:34 PM

 FIGURE 2

would undoubtedly be very high, but it

would not necessarily produce grain
growth.
At even higher overheating tempera-
tures reaching or exceeding 900 to 950°C
(1,740°F), the austenite grains may start
growing significantly. The higher the over-
heating temperature, the faster and the
larger the grains will grow. Upon cooling,
this very coarse, austenite grain structure
may give rise to a larger ferrite-pearlite
structure than the original one.
Only when the structure is coarser is
it possible to detect the occurrence of
grain growth by metallography. This was
the case that is illustrated in Figure 4,
when an intact steel beam was compared
with a badly warped one found during a
fire damage assessment.2-3 The degree of
softening was such (Figure 2) that the steel
could deform, buckle, distort, collapse,
bend, sag, or bulge, and its surface would
most likely be covered by a thick high-
temperature oxide scale that looked dark
blue to black. Metallography was not re-
ally needed in this case to assert that the
Metallography of partially transformed CS. steel had been overheated. It was sufficient
to see that it was badly deformed and
distorted and that it had a thick high-
temperature oxide scaling. The metallog-
raphy indicated overheating on the order
FIGURE 3 of 1,000°C (1,830°F).

The higher the

overheating
temperature, the
faster and the larger
the grains will grow.
If the austenite grains grow too large
and the cooling rate is rather fast, there
would not be enough nucleation sites and
time for the transformation to normal
ferrite and then pearlite to occur. This
process would cause the austenite to be-
come undercooled. At a small undercool-
Microstructure of partially transformed steel (original magnification 1,000X).
ing below the critical temperature, the
ferrite starts nucleating on the few existing
48 MATERIALS PERFORMANCE November 2004

NOV 2004 MP pp 1-51.indd 48 10/11/04 1:46:45 PM

 p g @@ f p g @@

FIGURE 4

(a) (b)
(a) Normal steel microstructure and (b) grain growth caused by overheating during a fire.

austenite grain boundaries and grows in FIGURE 5

a manner known as grain boundary al-
lotriomorphs.4-5 At larger undercooling
there is an increasing tendency for the
ferrite to grow from the grain boundaries
into the grains as plates—so-called Wid-
manstatten ferrite.
Figure 5 shows a sample removed from
a steel platform support subjected to fire
damage. The platform collapsed and was
badly warped. The steel had a thick, high-
temperature oxide layer. It was made of a
CS with a carbon content as low as that
in Figure 4, but the microstructure looked
quite different. It is a rather complex mi-
crostructure comprising light etching
constituents referred to as grain boundary
allotriomorph ferrite, Widmanstatten side
plates, and Widmanstatten intragranular
plates. In between the ferrite phase, the
dark etching constituent is fine pearlite.
So instead of seeing a normal ferrite pearl-
ite microstructure with a larger average
grain size, as compared to the original size Sample from a steel support overheated during a fire.
(as in Figure 4[b]), further grain growth

November 2004 MATERIALS PERFORMANCE 49

NOV 2004 MP pp 1-51.indd 49 10/11/04 1:46:55 PM

 caused by the cooling rate has produced flanges, piping, and even hot-rolled CS
a completely different and rather complex plates that have been finished at excessive
microstructure. temperatures and delivered in the as-hot
This type of CS microstructure is also worked condition. It immediately indicates
typically found in the coarse-grained zone that the steel had a rather coarse austenitic
adjacent to the fusion line in the heat- grain structure when it was finished and
affected zone of CS welds. The steel im- that it had not been normalized.
mediately adjacent to the weld is over-
heated to very high temperatures, causing
the transformation and subsequently ex-
Conclusion and Summary
CS is unique in the sense that it can
aggerated austenitic grain growth. The precipitate two phases that can have sev-
cooling rate in welds can occur rather fast, eral different microstructures—depend-
and the steel hardenability is greatly in- ing not only on the amount and presence
creased by the large austenite grains. of alloying elements, but also on its ther-
Manganese CS may even transform to mal history. Unless the transformation
bainite, martensite, or otherwise to Wid- processes are well understood, there is
manstatten ferrite and dark etching fine always the risk of misinterpreting its mi-
pearlite, in the coarse-grained zone that is crostructure. Such misinterpretation
adjacent to the weld fusion line. would prevent the realization of the full
This type of CS microstructure has also benefit gained by using metallography for
been seen in hot-forged pipe fittings, assessing the condition, evaluating metal-
lurgy, or investigating failures.

References
1. R.N. Parkins, Institution of Metallurgists, Modern
Metallurgical Texts, No. 5: Mechanical Treatment of
Metals (London, U.K.: George Allen and Unwin Ltd.,
1968), p. 298.
2. J.L. Hau, “Pressure Vessels Fire Damage Assessment
in a Refinery Unit,” CORROSION/92, paper no. 445
(Houston, TX: NACE, 1992).
3. J.L. Hau, “Assessment of Fire Damage to Pressure
Vessels in a Refinery Unit,” Corrosion 49, 5 (1993): p.
420.
4. D.A. Porter, K.E. Easterling, Phase Transforma-
tions in Metals and Alloys (New York, NY: Van Nostrand
Reinhold Co., 1981), p. 317.
5. L.E. Samuel, Optical Microscopy of Carbon Steels
(Metals Park, OH: ASM, 1980), p. 255.

JORGE HAU is a senior consultant at Capstone

Engineering Services, Inc., 1505 Hwy. 6 S., Ste.
250, Houston, TX 77077. He holds a mechani-
cal engineering degree from Venezuela and
M.Sc. and Ph.D. degrees in metallurgy from
The University of Birmingham in England.
He spent 7 years as a senior corrosion and
materials engineer in a petroleum refinery and
10 years as a researcher and consultant in a
petroleum research center in Venezuela, mostly
involved in refining. He has been involved in re-
search on degradation mechanisms for metallic
materials in oil-refining service and
has performed many failure analyses and
investigations.

50 MATERIALS PERFORMANCE November 2004

NOV 2004 MP pp 1-51.indd 50 10/11/04 1:47:04 PM

 p g @@ f p g @@

November 2004 MATERIALS PERFORMANCE 51

NOV 2004 MP pp 1-51.indd 51 10/11/04 1:47:19 PM